Bone marrow transplantation (BMT) is now the treatment of choice for a variety of hematological disorders, but the optimal marrow donor, an HLA identical sidling, is available for only 30% of the patients who might be cured by BMT. BMT using alternative donors (i.e., partially matched relatives or unrelated individuals) can be successful, but these transplants have higher rates of graft failure, increased incidence and severity of graft-versus-host disease (GvHD), and more frequent complications related to delayed or inadequate immune reconstitution. The long-term goal is to understand the molecular mechanisms of immunological recognition and to apply that knowledge to the development of effective therapeutic strategies to prevent or overcome life-threatening immunological problems such as GvHD and graft rejection. There are three major objectives related to definition of HLA disparity between marrow recipient and unrelated donors: (1) determine the extent of HLA disparity, (2) define the cellular recognition of the HLA disparities, and (3) explore the relationship between these variables and immunological complications following BMT.